When a fluid is delivered by a pump a number of problems can arise. Thus, at a constant speed of the motor driving the pump the volume flow can be undesirably modified by pressure fluctuations in the system, e.g. can fall when the plant pressure rises. In addition, a number of faults can arise, such as an inadmissibly high cavitation, a flat or shallow pressure rise at the start of the pressure stroke with a reduced discharge capacity due to air or gas inclusions, leaks to the outer chamber and relative to the inlet and outlet valve (suction/pressure valve). If such faults occur, they should either be eliminated or there should at least be a corresponding fault report.
For fault detection purposes a system is known, in which a pressure sensor is fitted to the hydraulic chamber of a pump and measures the pressure in the chamber. This pressure is not representative for the pressure in the dosing or metering chamber, particularly as a result of the inherent rigidity of the diaphragm, which applies to a greater extent with the double diaphragms used for leak detection and also due to a possibly inserted return spring, without which a reliable suction function is often no longer ensured. Criteria for the detection of fault functions should solely be the average suction and discharge pressure, as well as the pump efficiency, which can only describe a fault behavior in a very global manner.
It is also known to provide a pressure sensor on the pressure side and to generate fault signals if the system pressure drops below a lower preset value and exceeds an upper preset value. This neither permits a pump control, nor a precise fault detection and identification.
For the detection of imperfect operating conditions of a pump it is known to record in time-based manner the pressure distribution over the piston stroke, in that during the piston travel pressure values are continuously recorded at constant time intervals and the pressure change is investigated over time. It is disadvantageous that this presupposes a known, fixed piston or motor movement with a constant angular velocity.
Therefore the problem of the invention is to provide a method and a device of the aforementioned type, by means of which a reliable, precise fault detection and identification is possible.